Japan’s 2026 year-end and New Year holiday runs six consecutive days, from Tuesday, December 29, 2026 to Sunday, January 3, 2027. Only one weekday sits in front of it — Monday, December 28 — so anyone who takes that day off gets a nine-day break starting Saturday, December 26. Bookings typically get going in earnest from late September into October, which means the deadline for pricing and plan design is already here.

In this article we use MetroEngines Research data to examine room rates for this year-end/New Year period by area and property type across Japan. The short version: the peak rate falls on December 31 for city hotels and for onsen/resort properties alike. What differs is not the height of the peak but the width of the peak. For Tokyo business hotels, the peak lasts four days, December 28–31, and by January 2 rates drop into a trough more than 20% below normal. Shizuoka ryokan and resort hotels, by contrast, hold a plateau at least 15% above normal for eight days, December 27 through January 3, and stay above normal-month levels into mid-January. The period everyone calls the “year-end/New Year selling season” is nearly twice as long for some property types as for others.

Monthly ADR doesn’t show that year-end/New Year is expensive

We start with monthly estimated settled ADR — and immediately hit the first trap. Compare monthly ADR for December 2026 and January 2027 with the average of the most recent normal months, October and November 2026, and both business hotels and city hotels come in below normal months.

Taking the median of multiples calculated by prefecture and property type, business hotels come in at x0.96 for December and x0.91 for January, and city hotels at x0.93 and x0.90. Resort hotels, by contrast, are clearly higher at x1.05 and x1.10, and ryokan at x1.08 and x1.11. Running the same calculation across all 47 prefectures gives January multiples of x0.93 for business hotels and x0.89 for city hotels against x1.05 for resort hotels and x1.12 for ryokan — the same pattern, with city properties below 1.00 and onsen/resort properties above it. For city properties, December and January are not “expensive months” when viewed as monthly averages; if anything, they look like months that have fallen off from the autumn peak season.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

The reason is clear. Year-end/New Year demand lasts effectively six or seven days, from December 28 to January 3. The remaining 24 days of December and the remaining 24 days of January are when business demand in cities is at its weakest. A monthly average dilutes that short, tall peak with long, low troughs. If you set prices by looking only at monthly budgets and reports, the year-end/New Year window is structurally invisible.

Change the baseline and the picture changes too. With the January–November 2026 average set at 1.00, December estimated settled ADR is above the annual average almost everywhere: 1.21x in Beppu, 1.17x in Tochigi and Tokyo, and 1.16x in Hyogo, Hokkaido, Fukuoka and Aichi. Across all 47 prefectures, Okinawa (0.97x) is the only one where December falls below the annual average, and the biggest premiums are in regional prefectures such as Kagoshima (1.31x) and Fukushima and Tokushima (1.30x each). In other words, city properties only looked “below normal” because the baseline was the autumn peak of October and November. Given the seasonal pattern in which Tokyo’s listed prices hit their annual high in November, the conclusion flips depending on which baseline you choose.

December and January rates in major areas — the premium over the annual average

Here are the levels by area. The table below shows estimated settled ADR for prefectures and major onsen towns, along with the multiple of the January–November 2026 average. For reference, it also shows the listed price for the same month (all-plan average, tax included). The gap between estimated settled ADR and listed price comes from unsold high-priced plans staying on the listings and from the share of meal-inclusive plans.

Table 1: Estimated settled ADR in major areas (December 2026 and January 2027) and multiple of annual average

Area	Dec est. settled ADR	Jan est. settled ADR	Dec / annual avg	Jan / annual avg	Dec listed price	N (Dec)
Hakone	¥29,400	¥29,300	1.10	1.10	¥65,900	162
Atami	¥26,500	¥26,800	1.12	1.13	¥61,900	97
Yufu	¥26,800	¥26,600	1.15	1.14	¥53,900	130
Karuizawa	¥24,100	¥24,200	0.99	1.00	¥62,000	48
Yugawara	¥22,700	¥21,300	1.14	1.07	¥49,400	50
Kusatsu	¥19,700	¥20,300	1.14	1.18	¥47,600	62
Beppu	¥18,800	¥18,800	1.21	1.21	¥39,700	115
Tokyo	¥17,500	¥13,800	1.17	0.92	¥31,200	1,034
Kyoto	¥17,300	¥13,700	1.04	0.82	¥39,100	540
Oita	¥16,900	¥17,200	1.14	1.16	¥40,400	410
Hyogo	¥15,700	¥16,100	1.16	1.19	¥45,200	475
Shizuoka	¥15,300	¥15,200	1.14	1.13	¥37,800	756
Nagano	¥14,300	¥14,500	1.15	1.17	¥32,500	646
Hokkaido	¥12,500	¥13,100	1.16	1.21	¥29,800	795
Okinawa	¥11,900	¥11,600	0.97	0.94	¥28,400	497
Osaka	¥10,800	¥10,300	1.12	1.07	¥18,600	618

N is the number of properties used to calculate estimated settled ADR. The annual average is the simple average of estimated settled ADR for January–November 2026. Source: MetroEngines Research; compiled by HotelBank Editorial Team

January comes in below the annual average in Tokyo and Kyoto (0.92x and 0.82x), while in Hokkaido January is even higher than December (1.21x). Six days of year-end/New Year demand can’t explain that gap. In Hokkaido’s case, all of January falls within the ski-season peak. In fact, the daily index discussed below shows Hokkaido more than 15% above normal on 16 of the 26 days from mid-December to mid-January — not a year-end/New Year peak but a plateau spanning the whole winter season.

Ski resorts such as Hakuba and Kutchan carry premiums of a completely different order: Kutchan’s estimated settled ADR for January reaches ¥58,200 (3.7x normal months). However, only 21 and 16 properties respectively feed the estimated settled ADR in these areas, and because only properties open in winter remain, the sample itself changes with the season. We treat them as reference figures rather than as points of comparison for rate levels.

Day by day, the peak starts on December 28 and tops out on December 31

Now to the core of the analysis: using daily data to pull out the peak that monthly averages hide. Lining up daily year-end/New Year prices as they are mixes in the ordinary weekend effect, so we compared a day-of-week adjusted index — each day’s listed price divided by the average for the same weekday over November 9–22, 2026. 1.00 corresponds to the normal level for that weekday.

N = 856+ Tokyo business hotels, 102+ Tokyo city hotels, 205+ Shizuoka ryokan, 102+ Shizuoka resort hotels (each day). Source: MetroEngines Research; compiled by HotelBank Editorial Team

All four series share two movements. The first is the step up from December 27 to December 28. Tokyo business hotels go from 1.04 to 1.24 and Shizuoka resort hotels from 1.21 to 1.47 — in every series this single day brings the biggest jump. December 28 is officially a weekday, but it is the first day off for people taking the nine-day break, and the market prices it as the start of the year-end/New Year block.

The second is that December 31 is the peak for every property type: 1.39 for Tokyo business hotels, 1.18 for Tokyo city hotels, 1.71 for Shizuoka ryokan and 1.90 for Shizuoka resort hotels. You sometimes hear that city rates soften at year-end and peak in early January, but at least on current listed prices, New Year’s Eve is the most expensive day in cities and onsen towns alike.

The width of the selling season nearly doubles by property type — 4 days for city properties, 8 for onsen

Because the peaks line up, the difference shows in the foothills. Count the days from December 26 onward on which the day-of-week adjusted index is 1.15 or higher — that is, at least 15% above normal — and the length of each property type’s selling season comes straight out.

Table 2: Day-of-week adjusted index by property type — peak day and the tail (Tokyo and Shizuoka)

Series	Peak day	Peak index	Jan 2	Jan 3	Jan 4–12 avg	Days with index ≥1.15
Shizuoka – Resort hotels	12/31	1.90	1.36	1.26	1.05	8 days
Shizuoka – Ryokan	12/31	1.71	1.33	1.22	1.10	8 days
Shizuoka – Business hotels	12/31	1.39	0.90	0.92	0.93	5 days
Tokyo – Business hotels	12/31	1.39	0.78	0.76	0.78	4 days
Tokyo – City hotels	12/31	1.18	0.92	0.82	0.81	1 day

The index is the ratio to the average listed price for the same weekday over November 9–22, 2026 (= 1.00). Day counts are out of the 26 days from December 18 to January 12. Source: MetroEngines Research; compiled by HotelBank Editorial Team

Tokyo business hotels drop 44% in just two days, from 1.39 on December 31 to 0.78 on January 2 — and they don’t recover. The January 4–12 average is 0.78, and rates stay at the lowest level of the whole observation period (December 18 to January 12). The year-end/New Year window is effectively four days long, with troughs on either side. A short window also means how those four days are designed goes a long way toward determining how December and January finish.

Shizuoka ryokan and resort hotels, by contrast, stay at 1.15 or above for eight straight days. They hold 1.33–1.36 on January 2 and 1.22–1.26 on January 3, and their average from January 4 onward is still 1.05–1.10, above normal months. Business hotels in the same prefecture end after five days and fall to 0.93 in January, so this reads as a difference of property type, not of area. Property types where the stay itself — overnight with dinner and breakfast — is the purpose of the trip treat the second half of the holiday differently from property types where guests stay as a side effect of some other purpose.

This structure is also consistent with differences in the demand mix: the share of foreign guests differs sharply by facility type, at 12.9% for ryokan versus 47.2% for city hotels. Year-end/New Year is a period led by domestic demand — trips home and family travel — and the better a property type can capture that demand, the wider its peak.

Onsen towns don’t peak on the same day either — Hakone on New Year’s Day, Atami on New Year’s Eve

Nor can onsen towns be lumped together as “peaking at the end of December.” Lining up daily indices for six major onsen towns shows the peak date varying by area.

N = 195+ properties in Hakone, 143+ in Atami, 169+ in Beppu, 92+ in Karuizawa, 68+ in Kusatsu, 44+ in Yugawara (each day). Source: MetroEngines Research; compiled by HotelBank Editorial Team

Table 3: Day-of-week adjusted index at major onsen towns (key dates)

Onsen town	Peak day	Peak index	12/28	12/31	1/1	1/3
Atami	12/31	1.89	1.51	1.89	1.71	1.08
Karuizawa	12/31	1.82	1.38	1.82	1.54	0.99
Kusatsu	12/31	1.60	1.47	1.60	1.55	1.14
Yugawara	1/1	1.47	1.34	1.35	1.47	1.00
Hakone	1/1	1.37	1.30	1.31	1.37	1.05
Beppu	12/29	1.29	1.29	1.22	1.15	0.89

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Atami, Karuizawa and Kusatsu peak on December 31, with large premiums of 1.60–1.89. By contrast, Hakone, Kanagawa’s flagship onsen town, and neighboring Yugawara peak a day later, on January 1, with relatively moderate premiums of 1.37–1.47. Hakone holds a plateau of 1.22–1.37 from December 28 to January 2, with no concentration of demand on any single day. It sits within about two hours of the Tokyo metropolitan area, and the data suggest it tends to be chosen as “one night somewhere during the holiday” rather than as the place to ring in the New Year.

Beppu peaks even earlier, at 1.29 on December 29, and is back down to 0.89 by January 3. Onsen towns in Kyushu show a smaller and shorter year-end/New Year premium than those in the Kanto and Chubu regions. Just as Kusatsu forms a high price tier within Gunma, demand concentrates differently even among places that all count as “onsen towns.”

The practical implication is clear. In plateau-type areas such as Hakone and Yugawara, designing December 28 through January 2 as a single rate band is unlikely to leave much on the table. But in areas where demand concentrates on one day, such as Atami and Karuizawa, pricing December 31 and January 3 the same leaves revenue opportunities in both directions: the upside on New Year’s Eve isn’t fully captured, and inventory is more likely to go unsold on the 3rd.

Year-over-year comparison — Hokkaido and Aichi up more than 20%, Kansai roughly flat

Judging the level requires a comparison with last year. We set estimated settled ADR for December 2026 and January 2027 against the final figures for the same months a year earlier (December 2025 and January 2026). Note that the bases differ: last year’s side is final results, while this year’s side is estimated from listings at the time of the survey. This year’s figures will move as new plans are added and prices are adjusted.

Year over year. Last year is on a final basis; this year is estimated from listings as of September 2026. Source: MetroEngines Research; compiled by HotelBank Editorial Team

Of the 12 prefectures in the chart, Hokkaido is highest at +32.8% for December and +26.9% for January. Aichi also shows strong growth at +27.1% for December and +27.6% for January, but much of that comes from a low base the previous year (¥9,100 in December 2025); in absolute terms, its December 2026 level of ¥11,500 is the second lowest of the 12 prefectures in the chart after Osaka (¥10,800). Widening the view to all 47 prefectures, regional prefectures whose previous-December level was low, around ¥7,000, line up with growth of around +40%, such as Tokushima (+43.4% for December) and Toyama (+43.2%). Large growth rates need to be read together with how low the previous year’s level was. Tokyo is up a steady +10.5% for December and +4.2% for January, and Fukuoka +11.4% and +8.0%.

Meanwhile, Osaka (−2.8% for December), Kyoto (−3.3%), Oita (−2.2%) and Shizuoka (−1.6%) are slightly below last year for December. In January, however, all of them are back above last year or roughly flat. The December decline in the two Kansai prefectures could reflect either current listings not yet having built up to last year’s level or last December’s level itself having been high. It is worth watching how things move from October onward, when listings fill out.

Supply — the rooms available for the peak are already set

With a demand peak this short, how much room does the supply side have to move? We check the data on new openings.

Source: MetroEngines Research & Consulting (based on confirmed OTA listings)

There is an important caveat in reading this chart. The bars from September 2026 onward are suddenly low not because supply is falling. OTA listings for new properties start appearing a few months before opening, and many properties are listed only after they open, so the latest and upcoming months are structurally undercounted. In 2025, which is fully observed, there were 114 openings in November and 168 in December, with monthly counts ranging from 100 to 250 throughout the year. The figures of 11 for November 2026 and 16 for December 2026 are what has been observed so far and will rise as more listings appear.

With that in mind, looking at the depth of supply in rooms gives a different picture. Of the 168 properties that opened in December 2025 (a fully observed month), 71 were vacation rentals with 95 rooms, 25 were business hotels with 2,275 rooms, 15 were resort hotels with 949 rooms, 15 were hostels with 221 rooms, and 13 were guesthouses with 197 rooms — 4,124 rooms in total. Vacation rentals make up over 40% of openings by count but only about 2% of rooms.

Table 4: New openings by month — properties and rooms (based on confirmed OTA listings)

Opening month	Properties	Rooms	Per property	Observation status
Dec 2025	168	4,124 rooms	24.5 rooms	Fully observed
Aug 2026	133	3,486 rooms	26.2 rooms	Largely observed
Sep 2026	43	2,805 rooms	65.2 rooms	Partially observed
Oct 2026	25	1,418 rooms	56.7 rooms	Partially observed
Nov 2026	11	666 rooms	60.5 rooms	Partially observed
Dec 2026	16	1,550 rooms	96.9 rooms	Partially observed

Source: MetroEngines Research & Consulting (based on confirmed OTA listings)

Some perspective helps. Of the properties MetroEngines Research tracks, 27,162 properties with 1,266,938 rooms are confirmed as operating on OTAs. The 4,124 rooms that opened in December 2025 (fully observed) equal 0.33% of that room stock. Even if November and December 2026 openings eventually build up to last year’s level, the nationwide room stock will change by less than 1% ahead of the six-day year-end/New Year period.

Looking further ahead at the pipeline, accommodation facilities (50 rooms or more) being planned or built according to the Ministry of Land, Infrastructure, Transport and Tourism’s Building Construction Statistics number 18 projects with 3,815 rooms for 2027 and 8 projects with 1,428 rooms for 2028. These are based on building confirmation applications, however, and the counts are expected to rise as more applications are filed, so they should be read as the lower bound of the currently confirmed pipeline.

In short, room stock is effectively fixed for the year-end/New Year demand period. What can move is price, plan mix and inventory allocation.

Three points to check in September and October

Let’s turn the data into pricing practice. None of these points is about rejecting your current design; they are checks for drawing additional revenue out of a short window.

First, how many days do you define as year-end/New Year? The official holiday runs six days, December 29 to January 3, but prices start rising on December 28. For Tokyo business hotels, the index jumps 20 points in one day, from 1.04 on December 27 to 1.24 on the 28th. Leaving December 28 at the regular rate means missing the first night of the nine-day-break crowd. Conversely, at resorts and ryokan the index is still 1.22–1.36 on January 2 and 3, so a design that drops back to regular rates early leaves upside on the table.

Second, do you step your prices between the peak and the tail? In Atami, the index is 1.89 on December 31 but 1.08 on January 3 — a 1.75x spread within the same holiday period. In Hakone, it is 1.30 on December 28 and 1.37 on January 1, with little variation across the period. Setting a single rate for the whole period in a concentrated area like the former means losing revenue at both ends: not capturing everything on the peak day, and being too expensive and left unsold on the later days. It is worth checking which type your own property’s daily curve is closer to.

Third, how will you fill the trough after January 4? Tokyo business hotels average an index of 0.78 for January 4–12, the lowest level of the observation period. Hokkaido, meanwhile, holds 1.20 over the same period, and Shizuoka ryokan stay at 1.10. For city business and city hotels, early-to-mid January is a period that dilutes the revenue earned in the four year-end/New Year days. Whether you have separate measures aimed at that trough — multi-night discounts, long-stay plans, or corporate offers built around the post-New Year lull — will change how January finishes.

Right before year-end/New Year there are also separate demand peaks: Christmas (December 24–27) and the bonenkai (year-end party) season. Designing the year-end/New Year block together with the daily Christmas curve — when Tokyo rates run about 29.5% above a normal Friday — and with the room to create high-rate days on December weekdays makes the overall structure of December easier to see.

Summary

The 2026–27 year-end/New Year holiday runs six consecutive days, December 29 to January 3, or nine days for anyone who takes December 28 off. In monthly estimated settled ADR, city property types come in below the most recent normal months for both December and January, and year-end/New Year demand is buried in the monthly average. But pull the days out individually and remove the day-of-week effect, and a clear structure emerges: prices rise on December 28 and peak on December 31 for every property type.

The difference lies not in the peak but in the tail. Tokyo business hotels have only four days more than 15% above normal, fall to 0.78 by January 2 and stay there into mid-January. By the same measure, Shizuoka ryokan and resort hotels stay elevated for eight days and remain above normal months after January 4. Onsen towns split into types as well: Atami, Karuizawa and Kusatsu concentrate sharply on December 31; Hakone and Yugawara form a gentler plateau peaking on January 1; and Beppu is an early type that peaks on December 29.

On the supply side, the 4,124 rooms that opened in December 2025 (fully observed) are only 0.33% of the nationwide stock of 1,266,938 rooms. The stock barely moves ahead of the demand period. How you fit price and plan mix to a peak that is short and varies in width by property type — that is what the September-to-October design largely decides for the year-end/New Year season.

Related Articles

• Kanagawa Hotel Market 2026: Four Tiers and a ¥15,400 Autumn ADR Gap